Large multi-institutional two-arm randomized clinical trials, in which the primary endpoint is the time until a specific event (such as death from cancer) occurs, are generally conducted as fixed sample studies. These are, however, ethical and economic pressures to monitor the accumulating data and to terminate such studies earlier than was originally planned, when strong evidence emerges that one treatment is clearly superior to the other, or that both treatments are equally ineffective. Although theory and methods for such sequential monitoring have been available since 1940, they were not applicable to typical chronic disease trials, since the mathematical results were only valid if the results of the experiment could be observed instantly and the data were monitored continuously. Since 1975, major technical advances have appeared in the statistical literature that make it mathematically feasible to design sequential medical trials even when the response is delayed, and the data are monitored in groups. Increasing pressure is therefore being exerted on statistical units to apply these new techniques to on-going and future studies. A major obstacle is the total absence of a flexible and friendly software package to guide statisticians into selecting the right sequential design for their needs. We intent to develop such a microcomputer based software package with the help of consultants who have themselves made major contributions to the new statistical methodology for sequential clinical trials.